Jump to content

  •  

CNers have asked about a donation box for Cloudy Nights over the years, so here you go. Donation is not required by any means, so please enjoy your stay.

Photo

Canon Banding Help

  • Please log in to reply
80 replies to this topic

#1 HillTop

HillTop

    Ranger 4

  • -----
  • topic starter
  • Posts: 324
  • Joined: 26 Apr 2014
  • Loc: Belpre, OH

Posted 21 September 2015 - 09:01 AM

Anyone know a good way to remove wide horizontal banding?  I have a Canon 450D which is getting very annoying in that respect.  I have tried various camera settings and cooling and it looks like I'll be stuck with fixing them in post-processing.  I would think that dark frame subtraction would help as they show up in the stretched dark frames but so far that's a no-go (at least in DSS.)



#2 dugpatrick

dugpatrick

    Mariner 2

  • *****
  • Posts: 219
  • Joined: 13 Oct 2010
  • Loc: Round Rock, TX

Posted 21 September 2015 - 09:43 AM

I use a 450d and I haven't noticed any banding, except when I calibrate incorrectly.  It's important to match the ISO across the set of frames (bias/dark/flat/light).  I've noticed that while bias frames look similar between ISO's, the calibrated image will have banding if there was a mismatch.  If you've trying to recover from a situation where the flats ISO doesn't match the lights ISO, then I would skip the bias frames.

 

Doug


Edited by dugpatrick, 21 September 2015 - 09:43 AM.


#3 HillTop

HillTop

    Ranger 4

  • -----
  • topic starter
  • Posts: 324
  • Joined: 26 Apr 2014
  • Loc: Belpre, OH

Posted 21 September 2015 - 09:59 AM

Thanks dugpatrick, no problems there - always match up the ISO.  Same problem with or without flats and I usually don't use bias frames as they seem to mess with the dark subtraction in DSS, but I haven't taken the time to figure out why.



#4 Larry Phillips

Larry Phillips

    Explorer 1

  • -----
  • Posts: 50
  • Joined: 01 Oct 2007
  • Loc: Tucson, AZ

Posted 21 September 2015 - 10:04 AM

PixInsight has a built-in script for removing Canon banding.

 

Larry


  • okiedrifter likes this

#5 *skyguy*

*skyguy*

    Skylab

  • *****
  • Posts: 4,054
  • Joined: 31 Dec 2008
  • Loc: Western New York

Posted 21 September 2015 - 10:12 AM

If you have a copy of Photoshop and add the Astronomy Tools Actions Set by ProDigital ($20USD), you'll get a vertical and horizontal banding removal action that does a perfect job dealing with banding.


  • skycamper likes this

#6 dugpatrick

dugpatrick

    Mariner 2

  • *****
  • Posts: 219
  • Joined: 13 Oct 2010
  • Loc: Round Rock, TX

Posted 21 September 2015 - 11:28 AM

If you have a copy of Photoshop and add the Astronomy Tools Actions Set by ProDigital ($20USD), you'll get a vertical and horizontal banding removal action that does a perfect job dealing with banding.

+1 for the value of Astronomy Tools plugins

 

But, in my experience the 450d does not have banding.  Something is not right.

 

Doug



#7 Tonk

Tonk

    Cosmos

  • *****
  • Posts: 9,418
  • Joined: 19 Aug 2004
  • Loc: Leeds, UK, 54N

Posted 21 September 2015 - 11:40 AM

I have two 450Ds and neither exhibit banding so either you have a rogue 450D or your processing has an issue - so I agree with Doug



#8 HillTop

HillTop

    Ranger 4

  • -----
  • topic starter
  • Posts: 324
  • Joined: 26 Apr 2014
  • Loc: Belpre, OH

Posted 21 September 2015 - 11:44 AM

Been a while since I tried PixInsight, can't honestly remember if the banding tool helped much but I'm thinking it was marginal.  Plan on buying it but have been sinking too much money in my new observatory and haven't had much time for imaging.  As for photoshop - it's a no-go - software with a monthly fee is not gonna happen here.  Startools has a banding removal feature but I find that it either doesn't do much of anything or makes a big mess.  I need to dig some examples up to post - the bands I'm getting are quite wide and seem to have a color component.

 

Somewhat noticeable here - http://astrob.in/212748/0/

of course this has been processed to hide it to a degree


Edited by HillTop, 21 September 2015 - 12:08 PM.


#9 CP Kuiper

CP Kuiper

    Surveyor 1

  • *****
  • Posts: 1,836
  • Joined: 17 Jun 2015
  • Loc: Central Iowa

Posted 21 September 2015 - 12:27 PM

With the $9.99 monthly Adobe Photoshop Photography subscription we get both Photoshop CC 2015 and Photoshop Lightroom CC 2015.

As soon as upgrades are available we can update our software at no additional charge.

 

Back when we had to wait up to 24 months for the next release to get the new goodies, and pay $199 just to upgrade Photoshop. Lightroom was an additional upgrade cost.

 

So for the cost of 2 or 3 of fancy coffees or energy drinks we can have the most current industry standard photograph editing software available.

 

I spend a little less now to stay current than I did before, and now I don't spend in $199 chunks of money. 


  • Tonk, KevinUK and xiando like this

#10 HillTop

HillTop

    Ranger 4

  • -----
  • topic starter
  • Posts: 324
  • Joined: 26 Apr 2014
  • Loc: Belpre, OH

Posted 21 September 2015 - 01:15 PM

CP Kuiper - it's more the principle than the practical.  I want to pay for something and be done with it.  I hate insurance too.  :lol:


  • t_image likes this

#11 dugpatrick

dugpatrick

    Mariner 2

  • *****
  • Posts: 219
  • Joined: 13 Oct 2010
  • Loc: Round Rock, TX

Posted 21 September 2015 - 07:17 PM

Thanks dugpatrick, no problems there - always match up the ISO.  Same problem with or without flats and I usually don't use bias frames as they seem to mess with the dark subtraction in DSS, but I haven't taken the time to figure out why.

My suggestion: Try calibrating with bias/flat/dark/light frames, all at the same ISO.  With bias frames you want to have about 50 frames.  I usually check the box "Dark Optimization" on the Dark tab), but that's because my dark frames are rarely at the same temperature as my light frames.  I keep a library of dark frames, but it's still hard to match temperatures exactly when the sensor temperature drifts over the course of the imaging session.

 

I also dither when guiding, which will mask issues from differential flexure, and might possibly mask calibration issues too.

 

Doug



#12 xiando

xiando

    Fly Me to the Moon

  • *****
  • Posts: 6,692
  • Joined: 27 May 2015
  • Loc: Cloudy NEOhio

Posted 21 September 2015 - 07:22 PM

CP, yeah...I agree. I'm not terribly fond of the idea of subscription-based software in general, but in this case, it's fine with me. I can always fall back on my old copy of PS6 if I decide to cancel my subscription (but boy will I miss shake reduction....what a freaking awesome tool)


Edited by xiando, 21 September 2015 - 07:23 PM.


#13 HillTop

HillTop

    Ranger 4

  • -----
  • topic starter
  • Posts: 324
  • Joined: 26 Apr 2014
  • Loc: Belpre, OH

Posted 22 September 2015 - 01:09 PM

Went ahead and ordered another used camera.  Mine is unmodded and the new one is full-spectrum (no replacement glass.)  Will be interesting to compare the two as I've also read reports of the mod helping with noise in the red channel - something else I've struggled with at times.  Anyhow I'll be able to see if the banding is in the camera or not.



#14 HillTop

HillTop

    Ranger 4

  • -----
  • topic starter
  • Posts: 324
  • Joined: 26 Apr 2014
  • Loc: Belpre, OH

Posted 08 October 2015 - 09:52 AM

Same issue with new camera, pattern is different but similar.  It's not too bad in images unless I push them really hard (like galaxy shots), so maybe I'm just expecting too much.  



#15 Campos

Campos

    Apollo

  • *****
  • Posts: 1,266
  • Joined: 26 Dec 2012
  • Loc: Portugal

Posted 08 October 2015 - 11:18 AM

Hi,

 

If you are using an AC adapter to power the cameraI I sugest checking that out, have you tried with the battery or just the AC adaptor?



#16 HillTop

HillTop

    Ranger 4

  • -----
  • topic starter
  • Posts: 324
  • Joined: 26 Apr 2014
  • Loc: Belpre, OH

Posted 08 October 2015 - 11:47 AM

Campos - tried with and without battery, with and without SD card, computer controlled and wired remote - all have the same result.  



#17 sharkmelley

sharkmelley

    Aurora

  • *****
  • Posts: 4,598
  • Joined: 19 Feb 2013
  • Loc: UK

Posted 08 October 2015 - 02:38 PM

It might help if you post an example of what you are talking about.  I get banding on all my Canons but it may not be the same thing you are describing.  Banding is quite normal on Canons when you perform stacking and severe stretching - that's why there are so many tools for banding removal.

 

Mark



#18 leveye

leveye

    Aurora

  • ***--
  • Posts: 4,813
  • Joined: 06 Feb 2013
  • Loc: Metro Detroit Mi.

Posted 08 October 2015 - 07:41 PM

NIK Dfine removes horizontal and vertical banding associated with Canon sensors. Works wonders.



#19 Jon Rista

Jon Rista

    ISS

  • *****
  • Posts: 24,818
  • Joined: 10 Jan 2014
  • Loc: Colorado

Posted 08 October 2015 - 10:01 PM

Anyone know a good way to remove wide horizontal banding?  I have a Canon 450D which is getting very annoying in that respect.  I have tried various camera settings and cooling and it looks like I'll be stuck with fixing them in post-processing.  I would think that dark frame subtraction would help as they show up in the stretched dark frames but so far that's a no-go (at least in DSS.)

 

There are two things you need to do. 

 

First, stop using DSS. ;) Something I discovered when I switched to using PI to integrate is that I generally stopped having problems with banding. The AHD demosaicing algorithm used by DSS seems to be similar to the AHD algorithm used by Lightroom, and both cause banding. AHD in general, with Canon raw files, seems to be a little prone to banding, but the DSS algorithm seems to be particularly bad. 

 

PixInsight uses a VNG demosaicing algorithm by default. I never see banding with that, unless my exposures are woefully inadequate (i.e. 15 second subs or less at a dark site). PI also supports AHD, however it's implementation seems to be better than DSS's, and I rarely see banding with it.

 

So, now second. You need longer exposures. You are seeing banding because some of your signal is not sufficiently swamping all the noise in the camera itself. I use Canon cameras, and when I first started, when I was using shorter exposures, I frequently had banding in my images. Once I started getting longer subs, and was fully swamping read noise right down into the darkest background sky with skyfog, I stopped having problems with banding entirely, regardless of algorithm, when integrating with PI.

 

There are other ways to reduce banding, but they are all destructive to data in one way or another. The best solution is to eliminate the problem up front. So, get PixInsight (it has a 45 day trial period which you should use before buying it), which is invaluable as a processing tool for astrophotography in general, and use it to calibrate, register, and integrate. And use longer subs (your histogram should peak between 1/4 and 1/3 histogram). You should find that your banding issues disappear entirely.


  • Gucky likes this

#20 HillTop

HillTop

    Ranger 4

  • -----
  • topic starter
  • Posts: 324
  • Joined: 26 Apr 2014
  • Loc: Belpre, OH

Posted 09 October 2015 - 12:03 PM

Jon - used my Pixinsight trial already but never got around to learning the integration side of it.  I do plan on getting it, just too many irons in the fire.  Here's a highly stretched 420 sec dark frame at ISO 400 with the modded camera.  I haven't noticed the banding so much in my last couple outings since switching to the modded camera and using the LPS filter.  This has enabled me get 600 sec subs and could possibly go a little longer (at ISO 400.)  Like your idea of getting the skyfog level high enough to drown out the noise - I think I've read that but it never sunk in.  Thanks all for the responses.

Attached Thumbnails

  • 400Light001a.JPG


#21 HillTop

HillTop

    Ranger 4

  • -----
  • topic starter
  • Posts: 324
  • Joined: 26 Apr 2014
  • Loc: Belpre, OH

Posted 22 December 2015 - 04:46 PM

I need to do more testing, but my last image was integrated in DSS with 300 bias frames (vs. my usual 30 or so)  and came out without banding that I could detect.  The image also included a meridian flip without rotating the camera so that may have been an error on my part that contributed to the "fix" as I didn't manually rotate the images prior to DSS (don't know if it's necessary or not.)  Tried the batch process in PI for calibration/integration and got a hard lockup.  I'm assuming that PI is using all processing cores at a high capacity which would cause that as my rig is overclocked (yeah, I live on the edge.) :smashpc:   I'll try it out again after detuning a little.



#22 calypsob

calypsob

    Fly Me to the Moon

  • *****
  • Posts: 6,410
  • Joined: 20 Apr 2013
  • Loc: Virginia

Posted 22 December 2015 - 05:46 PM

Did you dither during capture?



#23 Jon Rista

Jon Rista

    ISS

  • *****
  • Posts: 24,818
  • Joined: 10 Jan 2014
  • Loc: Colorado

Posted 22 December 2015 - 07:34 PM

Jon - used my Pixinsight trial already but never got around to learning the integration side of it.  I do plan on getting it, just too many irons in the fire.  Here's a highly stretched 420 sec dark frame at ISO 400 with the modded camera.  I haven't noticed the banding so much in my last couple outings since switching to the modded camera and using the LPS filter.  This has enabled me get 600 sec subs and could possibly go a little longer (at ISO 400.)  Like your idea of getting the skyfog level high enough to drown out the noise - I think I've read that but it never sunk in.  Thanks all for the responses.


Glad it helps. I would avoid going much over 1/3rd histogram (back of camera), but if you can, I'd get as close to 1/3rd histogram as you can in each sub.

The noise in that frame looks quite typical of a Canon DSLR. They are pretty noisy suckers.

On a side note. Really wondering myself if the day will ever come when Canon's cameras have ultra low noise and ultra low dark current like their competition (not just Sony's Exmor, but Toshiba and Samsung sensors as well...they all have significantly better noise characteristics than Canon sensors. Samsung sensors are actually REALLY amazing...most neutral color I've ever seen, no strong red cast or anything like that (when link stretched, most other brands have some degree of red cast, Canon's being the worst)), VERY low read noise at higher ISO...less than 1e- once you get to ISO 1600. I wonder if Samsung is going to keep making cameras or not...they pulled the NX1 out of the European market (still selling elsewhere as far as I know), I was really hoping to see an upgrade to that camera soon here, as I think Samsung could be a solid competitor to Sony if they stick with it and execute properly.

#24 leveye

leveye

    Aurora

  • ***--
  • Posts: 4,813
  • Joined: 06 Feb 2013
  • Loc: Metro Detroit Mi.

Posted 22 December 2015 - 08:08 PM

NIK's Define filter plug-in for PS works wonders for vertical and horizontal banding.


  • calypsob likes this

#25 HillTop

HillTop

    Ranger 4

  • -----
  • topic starter
  • Posts: 324
  • Joined: 26 Apr 2014
  • Loc: Belpre, OH

Posted 23 December 2015 - 08:46 AM

Did you dither during capture?

Yes, but it never seems to make any difference in and of itself, even with a fairly large dither.  I get the feeling the problem lies more with the calibration frames than anything going on with the lights.

 

NIK's Define filter plug-in for PS works wonders for vertical and horizontal banding.

Don't have PS, not going there.  But thanks.

 

 

John, that's too bad if they drop out of the camera market, I read many good things about the new Samsung cameras from an AP perspective.  I want to move to a mono ccd myself, but observatory and new EQ-6 are gonna delay that purchase awhile. 




CNers have asked about a donation box for Cloudy Nights over the years, so here you go. Donation is not required by any means, so please enjoy your stay.


Recent Topics






Cloudy Nights LLC
Cloudy Nights Sponsor: Astronomics